Bingyu Zhang

Bingyu Zhang

Auto-generated profile, based on publicly available information

Bingyu Zhang

The Center for Health AI and Synthesis of Evidence (CHASE), University of Pennsylvania Philadelphia Pennsylvania USA

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 February 2024 in Annals of Internal Medicine
Verified
Journal Article
Closed access
QWQiong Wu
Qiong Wu
JTJiayi Tong
Jiayi Tong
BZBingyu Zhang
Bingyu Zhang
DZDazheng Zhang
Dazheng Zhang
JCJiajie Chen
Jiajie Chen
YCYong Chen
Yong Chen
The efficacy of the BNT162b2 vaccine in pediatrics was assessed by randomized trials before the Omicron variant's emergence. The long-term durability of vaccine protection in this population during the Omicron period remains limited. To assess the effectiveness of BNT162b2 in preventing infection and severe diseases with various strains of the SARS-CoV-2 virus in previously uninfected children and adolescents. Comparative effectiveness research accounting for underreported vaccination in 3 study cohorts: adolescents (12 to 20 years) during the Delta phase and children (5 to 11 years) and adolescents (12 to 20 years) during the Omicron phase. A national collaboration of pediatric health systems (PEDSnet). 77 392 adolescents (45 007 vaccinated) during the Delta phase and 111 539 children (50 398 vaccinated) and 56 080 adolescents (21 180 vaccinated) during the Omicron phase. First dose of the BNT162b2 vaccine versus no receipt of COVID-19 vaccine. Outcomes of interest include documented infection, COVID-19 illness severity, admission to an intensive care unit (ICU), and cardiac complications. The effectiveness was reported as (1-relative risk)*100, with confounders balanced via propensity score stratification. During the Delta period, the estimated effectiveness of the BNT162b2 vaccine was 98.4% (95% CI, 98.1% to 98.7%) against documented infection among adolescents, with no statistically significant waning after receipt of the first dose. An analysis of cardiac complications did not suggest a statistically significant difference between vaccinated and unvaccinated groups. During the Omicron period, the effectiveness against documented infection among children was estimated to be 74.3% (CI, 72.2% to 76.2%). Higher levels of effectiveness were seen against moderate or severe COVID-19 (75.5% [CI, 69.0% to 81.0%]) and ICU admission with COVID-19 (84.9% [CI, 64.8% to 93.5%]). Among adolescents, the effectiveness against documented Omicron infection was 85.5% (CI, 83.8% to 87.1%), with 84.8% (CI, 77.3% to 89.9%) against moderate or severe COVID-19, and 91.5% (CI, 69.5% to 97.6%) against ICU admission with COVID-19. The effectiveness of the BNT162b2 vaccine against the Omicron variant declined 4 months after the first dose and then stabilized. The analysis showed a lower risk for cardiac complications in the vaccinated group during the Omicron variant period. Observational study design and potentially undocumented infection. This study suggests that BNT162b2 was effective for various COVID-19-related outcomes in children and adolescents during the Delta and Omicron periods, and there is some evidence of waning effectiveness over time. National Institutes of Health.

30 March 2025 in medRxiv
Verified
Preprint
Open access
PDF file
BZBingyu Zhang
Bingyu Zhang
QWQiong Wu
Qiong Wu
RJRavi Jhaveri
Ravi Jhaveri
TZTing Zhou
Ting Zhou
MBMichael J. Becich
Michael J. Becich
YCYong Chen
Yong Chen
IMPORTANCE: Post-acute sequelae of SARS-CoV-2 infection (PASC) remains a major public health challenge. While previous studies have focused on characterizing PASC and identifying its subphenotypes in children and adolescents following an initial SARS-CoV-2 infection, the risks of PASC with Omicron-variant reinfections remain unclear. Using a real-world data approach, this study investigates the risks of PASC following reinfections during the Omicron phase in the pediatric population. OBJECTIVE: To investigate the risks of PASC diagnosis and 24 PASC symptoms and conditions after reinfection of SARS-CoV-2 during Omicron period in the pediatric population. DESIGN, SETTING, AND PARTICIPANTS: This retrospective cohort study used data from the RECOVER consortium comprising 40 children’s hospitals and health institutions in U.S. between January 2022 and October 2023. EXPOSURES: A second SARS-CoV-2 infection, confirmed by a positive polymerase-chain-reaction (PCR) or antigen tests, or a diagnose of COVID-19, occurring at least 60 days after the initial infection, compared to the initial infection. MAIN OUTCOMES AND MEASURES: PASC was identified using two approaches: (1) the ICD-10- CM diagnosis code U09.9 and (2) a symptom-based definition including 24 physician-identified symptoms and conditions. Absolute risks of incident PASC were reported, and relative risks (RRs) were calculated by comparing the second infection episode with the first infection episode groups using a modified Poisson regression model, adjusting for demographic, clinical, and healthcare utilization factors through exact matching and propensity scoring matching. RESULTS: A total of 465,717 individuals under 21 years old (mean [SD] age 8.17 [6.58] years; 52% male) were included. Compared to the first infection, a second infection was associated with significantly increased risk of an overall PASC diagnosis (RR, 2.08; 95% confidence interval [CI], 1.68-2.59), and with many specific conditions including: myocarditis (RR, 3.60; 95% CI, 1.46-8.86); changes in taste and smell (RR, 2.83; 95% CI, 1.41-5.67); thrombophlebitis and thromboembolism (RR, 2.28; 95% CI, 1.71-3.04); heart disease (RR, 1.96; 95% CI, 1.69 to 2.28); acute kidney injury (RR, 1.90; 95% CI, 1.38 to 2.61); fluid and electrolyte (RR, 1.89; 95% CI, 1.62 to 2.20); generalized pain (RR, 1.70; 95% CI, 1.48 to 1.95); arrhythmias (RR, 1.59; 95% CI, 1.45-1.74); abnormal liver enzyme (RR, 1.56; 95% CI, 1.24 to 1.96); fatigue and malaise (RR, 1.50; 95% CI, 1.38 to 1.64); musculoskeletal pain (RR, 1.45; 95% CI, 1.37 to 1.54); abdominal pain (RR, 1.42; 95% CI, 1.34 to 1.50); postural orthostatic tachycardia syndromes (POTS)/dysautonomia (RR, 1.35; 95% CI, 1.20 to 1.51); cognitive functions (RR, 1.32; 95% CI, 1.15 to 1.50); and respiratory signs and symptoms (RR, 1.29; 95% CI, 1.25 to 1.33). The risks were consistent across various organ systems, including cardiovascular, respiratory, gastrointestinal, neurological, and musculoskeletal systems. CONCLUSIONS AND RELEVANCE: Children and adolescents face significantly higher risk of various PASC outcomes after reinfection with SARS-CoV-2. These findings suggest a cumulative risk of PASC and highlight the urgent need for targeted prevention strategies to reduce reinfections, which includes an increased emphasis on initial or re-vaccination of children. Key Points: Question: Do children and adolescents face an increased risk of post-acute sequelae of SARS-CoV-2 infection (PASC) following reinfection during the Omicron era? Findings: During the post-acute phase, children and adolescents with reinfection are at statistically significant increased risk of incident PASC outcomes, including an overall PASC diagnosis and 24 most commonly complaints/symptoms/diagnoses associated with PASC. The risks remained consistent across different demographic and clinical subgroups. Meaning: These findings underscore the significant long-term health risks associated with SARS-CoV-2 reinfection in children and adolescents. Public health strategies should prioritize reinfection prevention, including enhanced vaccination efforts, to mitigate the burden of PASC in the pediatric population.

4 April 2025 in medRxiv
Verified
Preprint
Open access
PDF file
YSYishan Shen
Yishan Shen
BZBingyu Zhang
Bingyu Zhang
RJRavi Jhaveri
Ravi Jhaveri
JTJiayi Tong
Jiayi Tong
TZTing Zhou
Ting Zhou
TCthe RECOVER Consortium
the RECOVER Consortium
IMPORTANCE Nirmatrelvir-ritonavir, an efficacious treatment for acute COVID-19, has yielded conflicting evidence regarding its effectiveness in preventing Long COVID among adults. The effectiveness of nirmatrelvir-ritonavir in adolescents in real-world settings is not well understood. OBJECTIVE To evaluate the effectiveness of nirmatrelvir-ritonavir in adolescents aged 12 to 20 years given within the first five days of acute COVID-19 illness on illness severity during the acute (days 0 to 27) and post-acute phases (days 28 to 179). DESIGN, SETTING, AND PARTICIPANTS This target trial emulation used the electronic health records (EHR) data from twenty-nine medical institutions participating in the NIH’s RECOVER consortium. We included data collected between April 1, 2022 and December 31, 2023, and included all patients with a SARS-CoV-2 infection who were not hospitalized on the day of entry. Those who were treated with oral nirmatrelvir-ritonavir within 5 days after the positive test (n=2,923) and those who did not received nirmatrelvir-ritonavir within 5 days after the positive test during the acute phase of SARS-CoV-2 infection (control group, n=31,947) were identified. EXPOSURES Treatment with nirmatrelvir-ritonavir based on prescription records. MAIN OUTCOMES AND MEASURES We used modified Poisson regression models for binary outcomes to examine the effects of nirmatrelvir-ritonavir on acute phase hospitalization, emergency department (ED) visits, and outpatient visits as well as progression to moderate or severe acute illness among individuals treated with nirmatrelvir-ritonavir compared to control counterparts. Additionally, we assessed the medication’s associations with 16 conditions and symptoms during the acute phase of illness and its subsequent association with specific Long COVID diagnostic code U09.9 during the post-acute phase. We reported the absolute risks and estimated the relative risks (RRs), with adjustments made for demographic variables, clinical characteristics, and healthcare utilization, using propensity score matching. RESULTS The study cohort included 34,870 adolescents (median [Q1, Q3] age 16.0 [14.0, 18.0]); 46.0% male and 43.2% Non-Hispanic White. Among these patients, 2,923 were treated with nirmatrelvir-ritonavir within 5 days of cohort entry and 31,947 received no treatment. The absolute rates were 0.58% vs. 0.96% for hospitalization, 74.17% vs. 82.66% for outpatient visits, 1.81% vs. 2.28% for ED visits, 3.08% vs. 3.99% for moderate/severe acute illness, and 0.21% vs. 0.20% for long COVID diagnosis in the treatment vs. control group. After propensity score matching, during acute-phase period, compared to the control group, nirmatrelvir-ritonavir was associated with reduced risk of any hospitalization (relative risk (RR) 0.48; 95% confidence interval (CI) 0.29-0.80), outpatient visit (RR 0.86; 95% CI 0.82-0.90), and moderate/severe acute illness (RR 0.69; 95% CI 0.56-0.87). Additionally, nirmatrelvir-ritonavir was associated with reduced risk of 7 of 16 specified conditions in chest pain (RR 0.41; 95% CI 0.26-0.63), fatigue and malaise (RR 0.49; 95% CI 0.32-0.76), generalized pain (RR 0.65; 95% CI 0.45-0.95), headache (RR 0.60; 95% CI 0.44-0.83), mental health (RR 0.56; 95% CI 0.39-0.80), musculoskeletal (RR 0.63, 95% CI 0.43-0.94), and respiratory signs and symptoms (RR 0.61, 95% CI 0.52-0.71). No evidence was found to suggest that the treatment was associated with the risk of long COVID diagnosis U09.9 (RR 0.96, 95% CI 0.40-2.32). CONCLUSIONS AND RELEVANCE This study found that treatment with nirmatrelvir-ritonavir in adolescents within 5 days of SARS-CoV-2 infection, was associated with reduced risk of acute-phase COVID-19 illness and subsequent healthcare utilization. However, treatment was not associated with a lower risk of Long COVID diagnosis. Question Does treatment with nirmatrelvir-ritonavir reduce the acute COVID-19 illness severity and the risk of developing Long COVID in adolescents aged 12 to 20 years? Findings In this target trial emulation study of 34,870 adolescents with SARS-CoV-2 infection who were not hospitalized on the cohort entry day, 2,923 who were treated with nirmatrelvir-ritonavir exhibited lower acute illness severity including reduced risk of 7 of 16 specified conditions and symptoms, as well as reduced risk of acute-phase hospitalization and outpatient visits, compared with 31,947 who received no treatment. The association with reduced risk of Long COVID was not observed in this study. Meaning In adolescents, nirmatrelvir-ritonavir appears to be effective at reducing the severity of acute COVID illness but was not associated with the risk of developing Long COVID.

13 July 2026 in medRxiv
Verified
Preprint
Open access
HWHuiyuan Wang
Huiyuan Wang
BZBingyu Zhang
Bingyu Zhang
YLYuqing Lei
Yuqing Lei
YLYiwen Lu
Yiwen Lu
DZDazheng Zhang
Dazheng Zhang
YCYong Chen
Yong Chen
Glucagon-like peptide-1 receptor agonists (GLP-1RAs) have been linked to heterogeneous, potentially pleiotropic effects across organ systems, motivating outcome-wide comparative risk profiling in real-world data. A central challenge in such analyses is residual bias that remains after adjustment for observed confounders, which can distort effect estimates and mis-calibrate uncertainty. We present distributional diagnosis and calibration (DC), which uses panels of negative control outcomes (NCOs) to diagnose residual bias and calibrate uncertainty. DC evaluates null behavior via p -value uniformity and empirical coverage across NCOs, and uses the empirical distribution of NCO effect estimates to calibrate confidence intervals for prespecified primary outcomes. DC is modular: it can wrap around commonly used causal inference methods and operates directly on summary statistics, supporting collaborative research under data-sharing constraints. Using electronic health records from a large U.S. clinical research network (152.7 million patients), we compared GLP-1RAs with sodium–glucose cotransporter 2 inhibitors across 15 prespecified outcomes spanning cardiovascular, mental health, and genitourinary domains using four causal estimators. Across outcomes and methods, DC diagnostics revealed substantial and method-dependent residual systematic error. DC calibration attenuated systematic error signals observed in negative controls and yielded more stable, better-calibrated estimates for clinical outcomes, supporting DC as a practical strategy to strengthen the credibility of outcome-wide real-world CER. Disclaimer: The contents are solely the responsibility of the authors and do not necessarily represent the official views of, or an endorsement by, Food and Drug Administration (FDA)/Department of Health and Human Services (HHS) or the U.S. Government.

4 June 2025 in medRxiv
Verified
Preprint
Open access
PDF file
MSMaxwell Salvatore
Maxwell Salvatore
BZBingyu Zhang
Bingyu Zhang
HTHuilin Tang
Huilin Tang
YLYiwen Lu
Yiwen Lu
DZDazheng Zhang
Dazheng Zhang
YCYong Chen
Yong Chen
Glucagon-like peptide-1 receptor agonists (GLP-1 RAs) are increasingly prescribed for type 2 diabetes (T2D) and weight management, yet their broad health impacts in real- world settings remain understudied. Using data from the All of Us Research Program (n=18,746), we conducted both intention-to-treat and per-protocol phenome-wide association studies comparing diagnoses following GLP-1 RA prescription, including semaglutide-specific analyses, to those following sodium-glucose cotransporter-2 inhibitor (SGLT2i) and dipeptidyl peptidase-4 inhibitors (DPP4i) prescriptions in individuals with T2D. GLP-1 RAs were associated with reduced risks of genitourinary and dental conditions relative to comparators, while semaglutide was linked to lower risks of cardiac arrhythmia, hyperglycemia, and chronic kidney disease. However, GLP- 1 RAs were also associated with increased risks of dysthymic disorder and vitamin D deficiency. Time-to-event analyses revealed modest delays in diagnosis for key outcomes. These findings underscore differences in downstream associations across second-line T2D therapies and highlight semaglutide’s distinct profile. Results may inform clinical decision-making and motivate further research on effectiveness, safety, and personalized prescribing.

26 March 2025 in medRxiv
Verified
Preprint
Open access
PDF file
BZBingyu Zhang
Bingyu Zhang
LLLu Li
Lu Li
YLYiwen Lu
Yiwen Lu
JGJingchuan Guo
Jingchuan Guo
JBJiang Bian
Jiang Bian
YCYong Chen
Yong Chen
IMPORTANCE Telemedicine usage surged during the COVID-19 pandemic, shaping how patients access healthcare services. Its sustained role in post-pandemic healthcare may uncover long-term trends and variations in utilization. OBJECTIVE To characterize telemedicine utilization patterns from 2019 to 2024 and identify patient characteristics associated with telemedicine use. DESIGN, SETTING, AND PARTICIPANTS This retrospective cohort study analyzed outpatient visits across five hospitals within the University of Pennsylvania Health System (Penn Medicine) from January 1, 2019, to September 30, 2024. MAIN OUTCOMES AND MEASURES The primary outcome was the proportion of visits conducted through telemedicine. Multivariable logistic regression models were employed to assess association between telemedicine use and patient characteristics including demographics, insurance type, patient portal use, and socioeconomic status. RESULTS The study included 46,149,734 visits among 2,248,341 patients. Following the declaration of the COVID-19 pandemic in March 2020, telemedicine surged from 1% to 17% of outpatient encounters by April 2020, stabilizing at 8-13% for the rest of the year. Usage declined in 2021 but remained at 4-6% from 2022 to 2024. In multivariable models, older adults were less likely to use telemedicine compared to those under 40 years (40-64 years: aOR, 0.67 [95% CI, 0.67-0.67]; ≥65 years: aOR, 0.46 [95% CI, 0.46-0.46]). Higher telemedicine use was observed among women (male: aOR, 0.91 [95% CI, 0.91-0.92]), unmarried individuals (aOR, 1.10, 95% CI, 1.10-1.11), patient portal users (aOR, 1.44 [95% CI, 1.43-1.45]), patients with fewer comorbidities (Charlson Comorbidity Index scores ≥3: aOR, 0.87 [95% CI, 0.87-0.88]), those living farther from the place of service (5-15 miles: aOR, 1.04 [95% CI, 1.04-1.04]; ≥15 miles: aOR, 1.44 [95% CI, 1.43-1.44]; reference: <5 miles), lower-income individuals (<$50,000: aOR, 1.06 [95% CI, 1.06-1.07]; ≥$100,000: aOR, 0.91 [95% CI, 0.91-0.92]; reference: $50,000-$100,000), and primary care compared to specialty care (aOR, 1.19 [95% CI, 1.18-1.20]). Return patients used telemedicine more than new patients (new: aOR, 0.47 [95% CI, 0.47-0.47]). Telemdicine use varied by race/ethnicity, with lower use among Non-Hispanic Black (aOR, 0.89 [95% CI, 0.88-0.89]), Hispanic (aOR, 0.95 [95% CI, 0.95-0.96]), and Asian (aOR, 0.83 [95% CI, 0.82-0.83]) patients compared to Non-Hispanic White patients. Patterns varied across visit types (e.g., diabetes, mental disorders, sleep disorders, heart failure, COPD, CAD, and GI disorders), though younger, female, and geographically distant patients consistently used telemedicine more. Non-Hispanic White patients with mental disorders exhibited disproportionately higher telemedicine use, underscoring racial/ethnic differences that persisted during and after the pandemic, likely influenced by differences in access and coverage. CONCLUSIONS AND RELEVANCE Telemedicine use is higher among tech-friendly populations, including, younger individuals, female, return patients, and those living farther from healthcare facilities. However, difference by age, socioeconomic status, and race/ethnicity persist, suggesting barriers in access, digital literacy, and coverage. Targeted policies are needed to ensure equitable telemedicine adoption and accessibility for all patients. Question What has been the trajectory of telemedicine usage since the COVID-19 pandemic, and which sociodemographic factors are associated with its persistent use? Findings Telemedicine visits to the University of Pennsylvania Health System (Penn Medicine) surged with the onset of the COVID-19 pandemic in 2020. While telemedicine has since remained a significant component of outpatient care, its use has been uneven. Younger adults, women, patient portal users, unmarried individuals, return patients, and those with lower Charlson Comorbidity Index scores, were more likely to use telemedicine services. Moreover, racial and ethnic minorities, as well as socioeconomically disadvantaged populations, displayed varying patterns of telemedicine access, highlighting differences in utilization. Meaning Persistent differences in telemedicine use across primary care and specialty services highlight factors influencing its adoption, suggesting opportunities for targeted strategies to improve equity in digital healthcare access.